Omega 3/DataWarrior 4.7.3/smina version Apr. 29, 2017 with Vinardo scoring function/RF-Score VS v2
smina exhaustiveness = 16
We used Omega to generate 3D models of the ligands. DataWarrior was used to identify PDB files with the most similar ligands. Ligands were docked into the protein models from crystal structures with highest similarity ligands using smina with the Vinardo scoring function. The docked structures were minimized in smina and rescored using RF-Score VS.
"semi"-Crossdocking with Vinardo
Assumed pH 7.4
For this 20 pose tests, FragFP-DataWarrior was used to find the most structrually similar ligand bound to BACE-1 in the pdb databank. The most similar lignad-receptor pdb file was downloaded and processed with a local python script to pick out the receptor and one ligand which are further used in docking. Beside these receptor files, 4KE1.cif was checked against most of the 20 example, since it is a high resolution BACE-1 structure bound to a macrocyclic ligand and it is also has the same amino acid sequence with the sequence used for this contest. The intiail 3D models from the SMILES codes were generated with OpenEye's Omega.
exhaustiveness 16 (#default is 8)
Vinardo scoring function
Num_modes 5 #max number of poses to generate
autobox_add 2 or 4 #Amount of buffer space to add to auto-generated box, 4 as default. It was reduced to 2 when ligands are docked outsize the active site.
autobox_ligand #the position and size of the docking box was auto taken from the PDB's orginial ligand.
minimize # the result of each pose was minimized within smina
Docking runs were executed with the above specified parameters while default values
were applied for the rest of the variables. The top 5 poses from the Vina docking are submitted with this protocol. # The docking procedure requires the intial 3D pose files of the 20 ligands, and a PDB list of the each most relevant ligand bound PDB code and the 4KE1.cif. The docking procedure was performed with a local python script utlizing smina, which is a fork of Autodock Vina. And Vinardo scoring function was applied since it showed improved re-docking accuracy. The python script took the PDB code and fetch the cif file from PDB band, and then processed it into a receptor only pdbqt file, and the script also saved a PDB-ligand only pdbqt file which is used for autobox. Then the smina software was called with all the above-described input files and parameters. The result poses were visually checked using pymol, poses which are structurally and positionally similar the original ligands in the crystal structures are favored.